적응형 지도 대비 학습을 통한 효과적인 얼굴 위조탐지 방법
Design and Implementation of Deep Learning Network to Detect Various Face Spoofing Attack
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
딥러닝 기술의 발전으로 인하여 다양한 응용 분야에서 컴퓨터 비전 기술이 중요한 역할을 하고 있지만, 고성능 IT 기술의 발전으로 사진 및 영상의 위조가 쉬워짐에 따라 얼굴 인식 시스템의 보안에 큰 위협이 되고 있다. 본 연구는 생체인식 기술의 발전과 함께 증가하는 위조 기술에 효과적으로 대응하기 위하여, 지도 대비 학습(Supervised Contrastive Learning)을 활용한 얼굴 위조 탐지 방법을 제안하였다. 이방법은 데이터 불균형 문제를 효과적으로 해결하며, 다양한 얼굴 위조 공격에 구애받지 않고 얼굴의 진위여부를 효율적으로 판단할 수 있다. 실험 결과, 다양한 위조 방법에 대응하여 약 93%의 검출 정확도를 달성하였으며, 기존 연구 대비 최대 17%의 성능 향상을 보였다. 이러한 기술은 진위 여부를 판단하는 다른데이터셋에도 적용 가능할 뿐만 아니라, 새롭게 등장하는 위조 방법에도 효과적으로 대응할 수 있다.
Abstract
Advancements in deep learning technology have made computer vision techniques play a crucial role in various application fields. However, the development of high-performance IT technologies has made it easier to forge photos and videos, posing a significant threat to the security of facial recognition systems. This study proposes a facial forgery detection method using Supervised Contrastive Learning to effectively respond to the increasing forgery techniques alongside the advancement of biometric technologies. The proposed method effectively addresses the data imbalance issue and efficiently determines the authenticity of faces regardless of the type of facial forgery attack. Experimental results show a detection accuracy of approximately 93% against various forgery methods, achieving up to a 17% performance improvement compared to previous studies. This technology can be applied to other datasets for authenticity determination and can effectively respond to newly emerging forgery methods.
참고문헌 (25)
- "Quotable Quote" (2024)[URL]
- JD Trobe, "The Neurology of Vision", pp.100~104 (2001), Oxford Univ Press
- DA Forsyth, "Computer Vision: A Modern Approach", pp.85~93 (2021), Prentice Hall
- R Brinkmann, "The Art and Science of Digital Compositing", pp.1~14 (2008), Morgan Kaufmann
- C Chen, "Deep Learning for Cardiac Image Segmentation : A Review", Frontiers Cardiovascular Med, 7, pp.1~33 (2020)
- R Szeliski, "Computer Vision: Algorithms and Applications", pp.88~96 (2022), Springer Germany
- J Xu, "A Deep Learning Approach to Building an Intelligent Video Surveillance System", Multim Tools App, 80, pp.5495~5515 (2020)
- O Kasim, "Deep learning-based efficient and robust image forgery detection", Multim Tools App, 83, pp.59819~59838 (2024)
- KM Koppenhaver, "History of Forgery" (2007), Humana Press
- SR Ahmed, "Analysis Survey on Deepfake detection and Recognition with Convolutional Neural Networks", 2022 International Congress on Human-Computer Interaction, Optimization and Robotic Applications (HORA), pp.1~7 (2022)
- Y Tian, "Face Anti-Spoofing by Learning Polarization Cues in a Real-World Scenario", Proceedings of the 4th International Conference on Advances in Image Processing, pp.129~137 (2020)
- SK Singla, "Biometric System-Challenges and Future Trends", 8th International Conference on Computing for Sustainable Global Development, pp.647~651 (2021)
- Z Yu, "Benchmarking Joint Face Spoofing and Forgery Detection with Visual and Physiological Cues", IEEE Trans Dependable Secure Comput Early Access, pp.1~15 (2024)
- X. Lin, "Suppress and Rebalance: Towards Generalized Multi-Modal Face Anti-Spoofing", The 4th IEEE, 2024, ~
- X Guo, "Multi-domain Learning for Updating Face Anti-spoofing Models", European Conference on Computer Vision 2022, pp.230~249 (2022)
- A Jourabloo, "Face de-spoofing:anti-spoofing via noise modeling", European Conference on Computer Vision 2018, pp.297~315 (2018)
- Z Wang, "Deep Spatial Gradient and Temporal Depth Learning for Face Anti-Spoofing", IEEE, 2020, 5041~5050
- "What Is Liveness Detection and How It Works for Fraud Prevention" (2024)[URL]
- "Images for Face Recognition" (2020)[URL]
- JM Johnson, "Survey on Deep Learning with Class Imbalance", J Big Data, 6(27), pp.1~54 (2019)
- A. Kulkarni, "Foundations of Data Imbalance and Solutions for a Data Democracy", pp.83~106 (2020), Academic Press
- "Merriam-Webster Dictionary" (2024)[URL]
- T Chen, "Big Self-Supervised Models are Strong Semi-Supervised Learners", 34th Conference on Neural Information Processing Systems, Vancouver, pp.22243~22256 (2020)
- P Khosla, "Supervised Contrastive Learning", Proceedings of the 34th International Conference on Neural Information Processing Systems, pp.18661~18673 (2020)
- T Chen, "A Simple Framework for Contrastive Learning of Visual Representations", Proceedings of the 37th International Conference on Machine Learning, pp.1597~1607 (2020)